2022-08-20 14:57 UTC - we are investigating some jobs failing on internal error. Next update in 30 minutes.
2022-08-20 15:30 UTC - we see the internal error is caused by one of the internal components failing to call Azure API, however don't know the root cause. We restarted one of the instances and see no error for now. We continue monitoring the issue.
On 24th September 2022 between 10:00 CET and 10:30 CET, https://connection.eu-central-1.keboola.com will undergo planned maintenance during which one of our internal databases will be upgraded. As a result, any loads or unloads to storage will be paused during that time for up to 30 minutes. Any currently running jobs will continue to run. However, they may be delayed by up to 30 minutes as well.
2022-08-15 10:40 UTC - We are investigating component job delays in connection.keboola.com. Next update when new information will be available or in hour.
2022-08-15 11:14 UTC - We have identified the root cause of the problem and we are working on a solution. Next update when new information will be available or in hour.
2022-08-15 14:04 UTC - We've added another worker to help with the workload and the processing times has returned to normal
2022-08-15 15:32 UTC - We are investigating reoccurrence of the issue causing jobs stuck in Created state. Next update when new information will be available or in hour.
2022-08-15 15:59 UTC - The reoccurrence was resolved. We keep monitoring the situation closely, but at the moment job runtimes should be back to the normal.
2022-08-06 16:35 UTC - We are investigating higher error rates and latencies in connection.north-europe.azure.keboola.com stack. Next update in 30 minutes.
2022-08-06 17:20 UTC - Higher errors rates are resolved and latencies are back to normal. Schedules might be delayed, the backlog is processed.
2022-08-06 17:46 UTC - Backlogs is cleared. This incident has been resolved.
2022-08-04 09:40 UTC We are investigating bug, which causing error pages rendered as html source code.
2022-08-04 09:55 UTC Issue is now resolved all systems are working normally.
We are sorry for the inconvenience.
We are experiencing an issue with possibility of tables with _tmp suffix might be dropped.
If you manipulate with a table <X> and there also exist a table with name <X>_tmp, the table <X>_tmp could be accidentally dropped.
These tables (<X> and <X>_tmp) have to be in the same bucket and your project has to have Snowflake backend. We found the root cause and fix will be deployed in 2 hours.
Affected tables will be restored.
2022-08-02 17:00 UTC - We have deployed the fix of this issue and fixed affected tables. The issue is now resolved. We contact affected customers directly.
We are sorry for the inconvenience.
2022-07-27 11:12 UTC - Snowflake transformations does not respect selected warehouse size during input mapping which might lead to slower processing times. We are working on a fix which might be deployed in 4 hours.
2022-07-27 14:59 UTC - The fix was not yet deployed due to unexpected complications. We are working on it and the fix should be deployed in 4 hours.
2022-07-27 17:34 UTC - The fix was deployed. All Snowflake transformations are now using correct Warehouse size for execution. The issue is now resolved.
We are experiencing an issue with possibility of tables with _tmp suffix might be dropped. These tables have to be in the same bucket, and there have to exist a table with the same name, just without the _tmp suffix. We found the root cause and we are working on a fix which might be deployed in 4 hours.
2022-07-27 11:17 UTC - We have deployed the fix of this issue. Affected customers were contacted directly. We'll undrop affected tables.
2022-07-27 16:34 UTC - We have fixed affected tables. The issue is now resolved. We are sorry for the inconvenience.
Starting on August 8th we will resume migrations to the new Job Queue in our Azure North Europe stack.
This is a follow up to the previous round of migration, and all relevant information can be found in our previous post.
If you're not sure if your project is already on the new Job Queue, go to orchestrations in your project and if the url ends with "orchestrations-v2" then your project is already on the new Job Queue.
If you have any questions regarding the migration, please contact us via support@keboola.com
We apologize for any inconvenience.